Effect of cellulose nanocrystals on crystallization ...

Contents lists available at ScienceDirect

Polymer

journal homepage: www.elsevier.com/locate/polymer

Effect of cellulose nanocrystals on crystallization kinetics ofpolycaprolactone as probed by Rheo-Raman☆

Debjani Roya, Anthony P. Kotulaa, Bharath Natarajana,b, Jeffrey W. Gilmana, Douglas M. Foxc,Kalman B. Miglera,∗

aMaterial Measurement Laboratory, National Institute of Standards and Technology, Gaithersburg, MD, 20899, USAbDepartment of Physics, Georgetown University, Washington, DC, USAc Department of Chemistry, American University, Washington, DC, USA

H I G H L I G H T S

• Rheo-Raman captures growth of modulus and crystallinity during crystallization.

• CNCs strongly increase crystallization kinetics of PCL.

• CNCs decrease percolation threshold and induces anisotropic crystalline structures.

A R T I C L E I N F O

Keywords:CrystallizationRamanCellulose nanocrystals

A B S T R A C T

The development of biocompatible polymer nano-composites that enhance mechanical properties while main-taining thermoplastic processability is a longstanding goal in sustainable materials. When the matrix is semi-crystalline, the nanoparticles may induce significant changes to crystallization kinetics and morphology due totheir ability to act as nucleating agents. To fully model this behavior in a process line, an understanding of therelationship between crystallinity and modulus is required. Here, we introduce a scalable model system con-sisting of surface-compatibilized cellulose nanocrystals (CNC) dispersed into poly(ε-caprolactone) (PCL) andstudy the effects of nanoparticle concentration on isothermal crystallization kinetics. The dispersion is accom-plished by exchange of the Na+ of sulfated cellulose nanocrystals by tetra-butyl ammonium cations (Bu4N+)followed by melt mixing via twin-screw extrusion. Crystallization kinetics are measured through the recentlydeveloped rheo-Raman instrument which extracts the relationship between the growth of the transient me-chanical modulus and that of crystallinity. With extrusion and increasing CNC content, we find the expectedenhancement of crystallization rate, but we moreover find a significant change in the relative kinetics of increasein modulus versus crystallinity. We analyze this via generalized effective medium theory which allows com-putation of a critical percolation threshold ξc and discuss the results in terms of a change in nucleation densityand a change in the anisotropy of crystallization.

1. Introduction

The development of biocompatible polymer nanocomposites fromrenewable sources is under active investigation because the added na-noparticles offer the possibility of reinforcement of mechanical prop-erties of the matrix polymer. Less studied is the possible modification ofcrystallization kinetics that can occur with nanoparticles. Successfulimplementation of such novel materials then requires that in addition tooptimization of properties, the concomitant modifications to the

behavior in processing operations be understood and, if possible, beexploited.

Cellulose nanocrystals (CNCs) are an important class of renewablefiller for polymer nanocomposites [1–11] because their length to dia-meter ratio (∼20) results in low particle entanglement, which aids inprocessing, while still possessing a high Young's modulus (130 GPa-250 GPa) [12]; this represents an advantage over other larger-sizedcellulosic fillers [13]. Further utility of CNCs stems from their highertensile moduli, low density, low energy consumption for production,

https://doi.org/10.1016/j.polymer.2018.08.007Received 19 April 2018; Received in revised form 9 July 2018; Accepted 5 August 2018

☆Official contribution of the National Institute of Standards and Technology; not subject to copyright in the United States.∗ Corresponding author.E-mail address: [email protected] (K.B. Migler).

Polymer 153 (2018) 70–77

Available online 07 August 20180032-3861/ © 2018 Published by Elsevier Ltd.

T

high specific properties, modest abrasivity, biodegradability, and rela-tively reactive surface properties [14–16]. For the semi-crystallinematrix, we use polycaprolactone (PCL) as it is a biocompatible andbiodegradable polymer used in numerous applications including tissueengineering, drug delivery and additive manufacturing [17–19].

A ubiquitous issue in polymer nanocomposites is the dispersion ofthe nanoparticle in the polymer [1]. For CNCs, the hydrophilic natureof the polysaccharides causes irreversible agglomeration in nonpolarpolymer matrices due to the formation of hydrogen bonds between thenanoparticles. Further issues that hinder their use as reinforcing agentsare high moisture absorption and poor wettability [13,14]. Functiona-lization of the nanoparticle surface is necessary to avoid aggregation innonpolar matrices; for example, CNCs have been modified by surfacecoating or grafting [20]. Recently, Fox et al. demonstrated that re-placement of Na+ with imidazolium or phosphonium cations can beused to modify CNC surface energy. The CNCs were melt mixed intopolystyrene and the surface modification was shown to result in im-proved dispersion and improved resistance to moisture uptake [21].

Nanoparticles with sufficient dispersion in a polymer matrix havebeen shown to enhance the crystallization rate of semi-crystallinepolymers by acting as nucleating agents due to their high surface area[22,23]. In some cases, this is advantageous because nanoparticles caninduce specific crystalline forms that increase fracture toughness andoptical clarity [24–26]. Faster crystallization can decrease the manu-facturing time, for example in injection molding applications where thematerial must be sufficiently solidified prior to ejection from the mold.Prior work on PCL/clay nanocomposites investigated by differentialscanning calorimetry (DSC) and rheological measurements showed thatthe crystallization behavior and crystalline morphology are stronglyaffected by the presence of the clay particulates [27]. In a separatestudy, Siqueira et al. [13] studied different PCL/CNC nanocompositeswith several surface modifiers (PCL/nanowhiskers and PCL/micro-fibrilated cellulose) via DSC; the results showed that the differences inspecific surface area and surface chemistry of nanoparticles affected thecrystallization behavior of the polymer.

A critical aspect is how the increase of the modulus correlates withthe growth of crystallinity. Numerous models have been developed inthe past to try to relate the evolution of these two parameters [28–31]and often they depend on separate measurements of modulus andcrystallinity. However, these experiments considered only single com-ponent thermoplastics, and did not consider how the relationshipsmight be changed by nanofillers. Furthermore, the experimental en-deavor suffers from uncertainties because of the difficulty in main-taining identical thermal histories, geometries and surface propertiesbetween the rheological measurement and that of the crystallinity[32–35]. Recently, Kotula and coworkers developed a hybrid instru-ment – a rheo-Raman microscope - that simultaneously measures thekinetics of crystallinity via Raman spectroscopy and measures the ki-netics of moduli growth through mechanical rheometry [36]. This workis meant to highlight the applicability of the rheo-Raman microscope tocharacterize structural and conformational changes directly related tothe rheological response of the material. For PCL, the data was best fitby a generalized effective medium (GEM) model with two parameters,which is described in more detail later. An important parameter of themodel is the critical percolation threshold ξc, which describes the nor-malized crystallinity when the system becomes mechanically perco-lating. In neat PCL, it was found that ≅ξ 0.35c over a range of tem-peratures.

When nanoparticles are highly anisotropic, such as for nano-fibers,the polymer crystallization process can grow directly off the nano-fiber.For the case of multi-wall carbon nanotubes that induce nucleation inisotactic polypropylene, it was found that the nucleating CNT induces acrystalline layer around directly around the nanotube, and the aniso-tropy of the crystalline layer directly mirrors that of the CNT [37,38].Anisotropic polymer crystallization has also been induced by singlewall carbon nanotubes in solution solvent based crystallization of

polyethylene and Nylon [39]. Thus anisotropic nano-fillers can changethe crystallization kinetics in two fashions: first it increases the overallnucleation density, and hence nucleation rate of the polymer. Second, itcan cause highly-anisotropic crystalline domains to form, rather thanspherulites. The implications of these two effects on the rheology-crystallization kinetics will be explored in this work.

In this work, we examine how CNCs affect the crystallization ki-netics of PCL and how they impact the relationship between the growthof modulus and crystallinity. First, to prepare the dispersions of CNC inPCL, we use a modified form of the ion exchange approach of Fox et al.[21]. We assess the dispersion through optical microscopy, transmissionelectron microscopy (TEM) and rheology. We then employ the rheo-Raman microscope to perform simultaneous measurements of crystal-linity and modulus following an isothermal temperature quench [36].Consistent with existing literature, we find that the nanocrystallinematerial enhances the crystallization rate, indicating that the CNCs actas nucleating agents. The simultaneous measurements then allow us todirectly plot the rheological modulus as a function of crystallinity andwe show that the system can still be model by the generalized effectivemedium (GEM) theory, however the percolation threshold is sig-nificantly decreased because of extrusion and nanoparticle addition.

2. Materials and methods

Materials: We prepare two types of PCL nanocomposites each withthree different CNC mass fractions. The PCL used in this work has aweight-average molar mass of 112 kg/mol and was received in pelletform from Scientific Polymer Products, Inc. In the first type of nano-composite (unmodified), CNCs were obtained from the University ofMaine in freeze dried powder form. These were melt mixed in a twin-screw extruder at 95 °C at three different quantities; (1, 5 and 10)percent by mass. For the second type (modified), the CNCs were ob-tained from a 12.1% by mass aqueous slurry of Na-CNCs from theUniversity of Maine. The CNC surface was then modified with tertbutylammonium through the ion exchange method discussed above [21] todisrupt/minimize CNC-CNC interactions and improve dispersion in thePCL matrix. In this work we use the cationic surfactant, tetra butylammonium rather than imidazolium or phosphonium, to better matchthe aliphatic nature of PCL. The same concentrations in PCL are thenprepared via the identical twin-screw protocol. We also employ twodifferent 0% controls; the first is obtained by melt extruding the neatpellet PCL in the twin-screw extruder under the same conditions asutilized for the composite and is referred to as neat extruded PCL. Thesecond PCL control is simply the as received PCL pellets from thevendor and is referred to as neat pellet PCL. The comparison between thetwo controls allows us to isolate the effect of the extrusion process itselfon crystallization kinetics, which turns out to be significant.

Rheo-Raman and optical microscope: This experimental setup em-bodies the integration of a Raman microscope and rotational rheometercoupled through an optically transparent base. The detail of the in-strument is described elsewhere [36]. We use the instrument to mea-sure isothermal crystallization kinetics at 42 °C by simultaneous mea-surement of dynamic modulus, Raman spectroscopy and polarizedoptical microscopy. The Raman spectroscopy measurements in thispaper were performed using 532 nm laser light operating with 10mWpower at the sample. For the experiments presented here, the exposurecollection time was 5 s, and four sample exposures were averaged to-gether. The sample of thickness 300 μm is heated to 100 °C, graduallycooled at 10 °C/min to 52 °C, then cooled at a slower rate of 2 °C/min to42 °C and maintained isothermally to crystallize. The modulus wasmeasured during small-amplitude oscillatory shear using a fixed strainamplitude of 0.01 from 100 °C to 42 °C and then at a lower strain of0.004 at the crystallization temperature. The oscillation frequency was6.28 rad/s. For these measurements, the objective was focused ap-proximately 100 μm below the upper plate. The rheo-Raman micro-scope is capable of polarized optical microscopy in reflection mode.

D. Roy et al. Polymer 153 (2018) 70–77

71

However, since the magnification of the images is too small for clearimages, we instead integrate the intensity of the depolarized reflectedlight that reaches the camera CCD chip. This provides a sensitivemeasure for the onset of birefringent crystalline structures caused bycrystallization.

Rheology: Separate measurements of the rheology only were con-ducted in the melt state in an ARES G2 strain controlled rheometer.Frequency sweep measurements in the range 0.4–100 rad/sec wereperformed for the neat extruded PCL and the nanocomposite samples.The amplitude of the strain was 0.03 and we used a parallel plategeometry of 25mm diameter. The measurements were conducted atT= 105 °C because this is near the temperature at which annealing wascarried out and also near the temperature at which the CNCs weremixed with polycaprolactone in the twin screw extruder.

TEM and optical imaging: All samples were cryo-microtomed at−120 °C (well below the glass transition temperature of PCL, which is−60 °C) into 100 nm and 500 nm sections for TEM and optical imaging,respectively. Optical imaging was performed with an Olympus BX 51microscope operating in transmission mode. The cryo-microtomedsections were collected on a glass slide and imaged using a 50× ob-jective. TEM sections were collected on copper grids and imaging wasperformed in an FEI Titan microscope at an accelerating voltage of200 kV, with a zero-loss energy filter. This filtering has been shown toenhance density-based contrast variations in the sample [40]. To fur-ther enhance material contrast, a 40 μm objective aperture was em-ployed.

Polarized Optical measurements (POM) for crystallization: Thesample was cooled and placed between the quartz disks of a Linkamshear cell, heated to 100 °C, and compressed to a thickness of 100 μm.The isothermal crystallization experiments were performed at 42 °C.Imaging was performed in transmission mode using a 50× objective.

3. Results and discussion

Fig. 1a-d shows optical and TEM images of the 5% modified and 5%unmodified samples. We first compute the total area fraction, calcu-lated by binarizing the images (Fig. S1) and summing the area occupiedby larger aggregates observable in the optical images, and the smallerones that require TEM. In the TEM image analysis, we do not distin-guish between individualized CNCs and CNC aggregates. We find thetotal area fraction approximately agrees with the initial mass fraction ofCNCs in the polymer matrix, indicated that both imaging modalities arerequired to account for the CNC content. Note that the aggregates areanisotropic. The details of the analysis are discussed in the SI (disper-sion quantification).

As a measure of dispersion, the average area and the equivalentspherical diameters of the aggregates were calculated using the imagesfrom the two length scales. For the 5% by mass nanocomposites, theaverage area of the aggregates with modified CNCs was about 2 ordersof magnitude smaller than that of the unmodified CNC sample. Thisroughly translates to an order of magnitude difference in the equivalentspherical diameter (Table 1). Uncertainties represent one standard de-viation in calculated values from multiple replicants. Fig. 1e shows thepeak-normalized particle size distributions for the 5% modified andunmodified CNCs estimated by summing weighted distributions fromthe optical and TEM measurements. It is apparent that the surfacemodification results in a significant reduction in the most probableparticle size, but also note the broad distribution of sizes. A similartrend in aggregate dimension is observed in the 1% samples (Table 1).However, in the 10% CNC nanocomposites, the modified samples dis-play highly aggregated morphologies. Analysis of the 10% modifiedsamples reveals a significant degradation of the dispersion quality ingoing from 5% to 10% CNC by mass (Table 1). Images from all 1% and10% modified nanocomposites are shown in Figs. S2 and S3 respec-tively. The 10% unmodified sample is highly heterogeneous, renderingit difficult to obtain a representative volume fraction of the material for

statistically relevant quantification. We do not consider this samplefurther.

The melt rheology results show a reinforcement effect with in-creasing mass fraction of CNC. In Fig. 2 we plot the complex viscosityand find a modest increase of viscosity in going from neat PCL to 1%,and then a stronger increase from (1–5) %. This is expected for nano-particles with modest (∼20) length to diameter ratios and indicatessuccessful dispersion in the melt state. The lack of increase from (5–10)% then indicates that further increase of CNC beyond 5% does not in-crease the number of dispersed particles, a result that is anticipatedfrom the imaging data in Table 1 and Fig. S3.

We now turn attention to the crystallization kinetics, first examiningthe effect of CNC surface modification on the crystallization kineticsusing rheology alone and then the relative growth of crystallinity andmodulus via the rheo-Raman measurements. The enhancement inmodulus growth that is induced by the CNCs is measured via smallamplitude rheology, following the isothermal protocol described aboveis shown in Fig. 3. There are two trends evident in this plot. The first isthat for a given CNC loading, the upturn in the modulus growth occursearlier upon surface modification; for example, one can compare the5% modified against the 5% unmodified. This result is in accord withthe increase in surface area that was reported earlier; it demonstratesthat the surface-modification procedure not only enhances the meltrheology, but that with better dispersion of nanoparticles, there is anincrease in the crystallization kinetics because CNCs act as nucleatingagents. The second trend is the increase in modulus growth rate for agiven modification type with increasing CNC content, up to 5%. This istrue for both the modified and the unmodified CNCs. Fig. 3 also showsthe growth of modulus between the two pure PCL samples: the neatextruded PCL and the neat pellet PCL. Remarkably, a significant en-hancement in crystallization kinetics occurs when the sample has been

Fig. 1. Optical images of a) 5% CNC unmodified and b) 5% CNC modifiedsamples. TEM images of c) 5% CNC unmodified and d) 5% CNC modifiedsamples. The scales are an order of magnitude different for better comparison ofthe CNC aggregates in the two images e) Calculated particle size distributionsfor the modified and unmodified samples.

D. Roy et al. Polymer 153 (2018) 70–77

72

through the twin-screw extruder, though the overall kinetics still do notexceed those of the 1%. This result is likely an indication that nano-particulate, such as from metal particles or gelled polymer particles arecaused by the extrusion process itself, either by residual particles fromprevious extrusions or by nanoparticles that slough off the surface ofthe extruder during the process. We have determined from GPC mea-surements that there is negligible modification in the molar mass dis-tribution upon extrusion (Fig. S4), indicating that it is unlikely that thisobservation is the result of chemical degradation of the polymer. Thisresult is interesting from the general perspective of comparing 0%model crystallization studies with those that come from industriallyextruded samples. Care should be taken to know if there is any pollu-tion of the extruded samples by the extrusion process itself.

We briefly describe the Raman spectra of molten and semi-crystal-line states of neat extruded PCL and a PCL/CNC composite (Fig. 4). Thedifferences between the spectra with and without CNC for a giventemperature is negligible indicating that the CNC peaks are weak atthese mass loading compared to the PCL, though CNCs are known tohave spectral features in the range of (1000–1500) cm−1 [41]. Toquantify the mass fraction of the crystalline phase, we use the approachof Kotula et al. [42] where the basis spectra were determined for theC=O peak. A linear relationship was found between DSC determinedcrystallinity and the normalized peak area of the basis spectra thatcontains a peak at 1722 cm−1, so that =α βI I/c cr tot where αc is thecrystalline mass fraction, Icr is the Raman intensity of the basis spectraat 1722 cm−1, Itot is the total of that in the C=O region and β was foundto be 1.26. Fig. 4b shows the deconvolution of the Raman spectra in theC=O region into three basis spectra including the one with a peak at1722 cm−1. To understand the relative growth of modulus and crys-tallinity induced by the extrusion and the nanoparticles, we simulta-neously measure the modulus from rheology, the crystalline massfraction from Raman spectroscopy and the reflected cross polarizedlight during isothermal crystallization.

Fig. 5 shows the rheo-Raman results of the isothermal crystallizationof three modified samples (1, 5 and 10) % as well as the controls that donot contain CNCs (for clarity, we only show results from the modifiedCNCs). For the rheological and modulus measurements (Fig. 5a) wenote the trend identified previously regarding the relative increase incrystallization kinetics with increasing CNC content (Fig. 3). Themodulus value reported here is consistent with the measurements ofWang et al. [43]. We do not observe an increase in modulus with in-creasing CNC mass fraction; this is likely due to the effect being small atthese volume fractions and our use of a sub-optimal geometry to makesuch a measurement (parallel plate rheometry is not accurate comparedto standard DMA for measurement of crystalline modulus). Fig. 5bshows the crystalline mass fraction during the isothermal crystal-lization. However, the modified 10% nanocomposite does not show anincreased rate. This is likely due to the reduction in available nucleatingarea due to the significant degradation of dispersion quality (Fig. S3),despite the two-fold increase in the number of potential nucleatingcenters (Table 1). By simple inspection, the same trend regarding theincrease in crystallization rate can be gleaned from the Raman mea-surements as was seen previously for the modulus enhancements. Theαc starts at negligible values, then increases measurably at times ran-ging from 200 s to 600 s depending on the concentrations of CNCs. The

Table 1Average area of CNC aggregate Aagg, Effective Spherical Diameter Dagg and Area Fraction obtained from image analysis.

Samples Average area of CNC aggregate Aagg (nm2) Effective Spherical Diameter Dagg (nm) Area Fraction (%)

1% modified 5.22× 103± 4.17× 103 68 ± 29 1.64 ± 0.451% unmodified 2.20× 105± 6.67× 104 526 ± 81 1.78 ± 0.055% modified 1.57× 103± 5.26× 102 44 ± 7 4.26 ± 0.265% unmodified 2.01× 105± 6.80× 103 506 ± 93 5.24 ± 0.3710% modified 8.3×105±4.87× 105 890 ± 273 15.02 ± 0.6310% unmodified 16.11× 105± 13.63× 105 1204 ± 565 6.49 ± 0.41

Fig. 2. a) Magnitude of complex viscosity versus angular frequency for PCL andPCL/CNC nanocomposites b) storage modulus and loss modulus versus angularfrequency for PCL and PCL/CNC (5%) nanocomposite. All measurements aredone at 105 °C.

Fig. 3. Storage modulus of the modified and unmodified 5% CNC, 1% CNC,neat pellet PCL and neat extruded PCL. All measurements are done at 42 °C.

D. Roy et al. Polymer 153 (2018) 70–77

73

crystallinity grows to an average mass fraction of 0.42 ± 0.01. Aninteresting feature of this data concerns the upturn of the modulus ascompared to the upturn in the crystallinity for a given sample. For theneat pellet PCL, the two upturns occur at roughly the same time point;however, for all the extruded samples, the upturn in modulus curveoccurs well before that of the crystallinity. We return to this point infurther discussion.

Fig. 5c shows the simultaneous reflected mode polarized opticalmicroscopy measurements for the same five samples. At, t= 0 s, thereflected depolarized intensity is negligible for the two control PCLsamples since they are fully molten and non-birefringent, but as CNCsare added it becomes finite and increases with the mass fraction of CNC.This non-zero intensity at t= 0 s, before crystallization starts, is due tothe birefringent nature of the CNCs themselves. The changes in re-flected light intensity with time, rather than the magnitude of the in-tensity itself then serves as a marker for the crystallization. For all cases,the integrated pixel intensity is initially constant with time; the re-flection mode imaging showed no structural features under crossedpolarizers. In separate measurements, polarized transmission opticalmicroscopy at higher magnification with a thinner sample did not re-veal significant growth of spherulites, instead a non-distinct “grainy”

structure appeared during crystallization (Fig. S5). At approximately480 s for neat pellet and 360s for neat extruded PCL, the intensity in-creased due to the appearance of birefringent structures and then afterreaching the peak value, the intensity decreased as sample became in-creasingly turbid due to light scattering structures growing in the bulk[44]. It should be noted that the increase in average pixel intensityappears prior to any observable conformational change in Ramanspectra or modulus change in both the neat PCL and 1% CNC sample.Prior results have also shown that light scattering appears well beforecrystallinity in rheology and X-ray scattering measurements [45–47]. Inthe 5% and 10% samples there is only a decrease in intensity of thepolarized intensity; this is due to the birefringence and turbidity ofthese samples from the CNCs themselves. Interestingly, the time atwhich the birefringent structures change coincides with the upturn inrheology. It is unclear at this point that the effect stems from an opticalinteraction between the CNCs and the crystallizing sample, or if it isfrom a change in the nature of the crystallinity itself.

A straightforward method to extract the changes in the relationshipbetween the growth of modulus and that of crystallinity as a function ofnanoparticle content is to plot the times to reach 50% of the logarithmof final elastic modulus value and 50% of the final crystallinity value foreach condition. Fig. 6 shows this analysis extracted from the data inFig. 5a and b; note that both controls at 0% CNC are shown. First, thereis the decrease in the time scales as CNC content is increased, and thereis a decrease in time scale from the case of neat pellet PCL to neat

Fig. 4. a) Raman spectra of the 5% PCL/CNC nanocomposite (upper twocurves) and the neat pellet PCL (lower two curves) in the melt (90 °C) and semi-crystalline (42 °C) states. b) Raman spectra in the C=O region showing thedeconvolution of the curve peak into melt (amorphous and dipole-dipole) andcrystal basis spectra at 1733 cm−1 and in the 1722 cm−1 respectively.

Fig. 5. a) storage modulus b) Raman crystallinity c) Intensity profile for neat(pure and extruded) PCL and PCL/CNC (modified) composites, all measure-ments were carried at isothermal temp, 42 °C.

D. Roy et al. Polymer 153 (2018) 70–77

74

extruded PCL; this effect was noted in Fig. 5. A closer examination re-veals that the ratio of time scales for 50% growth, t t/α G

1/2 1/2, is near unity(1.05) for the neat pellet PCL whereas it is 1.26 for the 5% samples. Thisreflects that the nanoparticles cause the upturn in the modulus to occurbefore the upturn in crystallinity.

In Fig. 7 we construct a plot of the modulus as a function of thedegree of space filling of the crystalline domains, ξ, where ξ=ϕ/ϕ∞.Here we define ϕ as the crystalline volume fraction calculated from themass fraction (αc) obtained using Raman measurements (equation (1)):

=

+−

ϕ α

αc

cν α

ν(1m c

c

)(1)

Here νm and νc are the specific volume of the melt and crystal phases,respectively and have been reported in the literature: νm =9.1 × 10−7 m3/g and νc =8.5 × 10−7 m3/g [48]. As the modulus and crystal-linity data are sampled at different rates, the modulus data is inter-polated so that they can be effectively be reported at the same time. Forclarity, we show only the 0% controls and the 5%. The (1 and 10) %curves lie near that of the 5% data. We can observe the stark differencebetween the neat pellet PCL and the other curves, which is again a re-flection of the fact noted above that for the CNC containing samples andthe extruded control, the upturn in modulus occurs before that ofcrystallinity.

We can fit the data in Fig. 7 to the GEM model, which is based onthe elastic version of the Krieger- Dougherty equation to develop asuspension-based model across the percolation transition [48]. TheGEM equation utilized is shown in Equation (2):

−−

++

−

+=

∗ ∗

∗ ∗

∞∗ ∗

∞∗ ∗

ξG G

G A Gξ

G GG A G

(1 )( ) ( )

( ) ( )( ) ( )

( ) ( )0m

q q

mq q

q q

q q

1/ 1/

1/ 1/

1/ 1/

1/ 1/ (2)

Here, ∗Gm is the magnitude of the complex modulus at t= 0 duringisothermal crystallization, ∞G* is the magnitude of the modulus at whencrystallinity has filled the space, and ∗G is the magnitude of the complexmodulus at time t. q is the scaling exponent and A is defined as:

=−

Aξ

ξ1 c

c (3)

ξc is then the critical percolation threshold.We fit Equation (2) to A and the scaling exponent q as shown in

Table 2. We therefore define =∗G ξ[ ] q/ c an intrinsic modulus thatdescribes an increase in the modulus due to small addition of crystallinematerial. The values reported in Table 2 further show that for the ex-truded samples the intrinsic modulus ∗G[ ] is higher indicative of non-spherical domains as in agreement with lower percolation resulting inhigh aspect ratio of the crystalline domains [49]. The fitting results inFig. 7 show that the model provides an acceptable fit to the storagemodulus over the entire crystallization process. The critical percolationthreshold ξc only accounts for the melt and semi-crystalline phases ofthe polycaprolactone matrix. To calculate the critical percolation frac-tion of the composite material, we multiply the critical percolationfraction by the volume fraction of polycaprolactone,

′ =+ −

ξ ξc cx

x x(1 )

p

p pρCNC

ρp

where xp is the polymer mass fraction, ρp is the

polymer density in the melt state (1.1 kg/m3), and ρCNC is the CNCdensity (1.6 kg/m3). The critical percolation fraction of the composite ′ξcis reported in Fig. 8, where at zero volume fraction ′ =ξ ξc c. The primaryresult is that the percolation threshold drops from 0.33 for the neatpellet PCL to values of approximately 0.1 for the other samples (i.e. theextruded pellet PCL, and the CNC containing samples). There is a slightdecrease in the composite critical percolation fraction ′ξc at the 10%CNC loadings. However, the decrease is comparable to the experimentaluncertainty and so we do not wish to overinterpret that result. The

Fig. 6. Half-time crystallization from storage modulus (rheology), and Ramancrystallinity of neat pellet PCL (solid symbol) and extruded PCL and PCL/CNCcomposites (hollow symbol) at 42 °C.

Fig. 7. Representative fit for GEM percolation model for neat pellet PCL, neatextruded PCL and 5% PCL/CNC nanocomposite; all measurements done at42 °C.

Table 2Composite critical percolation concentration ( ′ξc ), exponent (q) and ∗G asfunction of concentration of CNCs.

Samples ξc′ q [G*]

neat pellet PCL 0.33 ± 0.04 2.11 ± 0.37 6.40 ± 1.37neat extruded PCL 0.11 ± 0.03 1.04 ± 0.08 9.45 ± 2.421% modified CNC 0.12 ± 0.03 0.79 ± 0.09 6.67 ± 1.685% modified CNC 0.09 ± 0.02 0.70 ± 0.07 7.77 ± 1.6210% modified CNC 0.1 ± 0.03 0.78 ± 0.17 8.21 ± 2.50

Fig. 8. Percolation transition conc. for neat pellet PCL, neat extruded PCL andPCL/CNC composites, pellet form of the neat PCL has higher percolation conc.than the extruded forms.

D. Roy et al. Polymer 153 (2018) 70–77

75

scaling exponent q is related to the sharpness of the fitting curve at thepoint of the critical percolation fraction; lower numbers indicate asharper transition. While the relative error in this fitted exponent islarge, there is a clear trend towards lower values for the neat extrudedand then the CNC containing samples compared to the neat pellet. Thisis another indication of the change in crystalline kinetics, as manifestthrough the crystalline-modulus relationship with the addition of par-ticulates and CNCs. The values ′ξc and q are reported in Table 2.

We can thus consider two distinct trends that are embedded in thedata: changes of crystallization rate and of percolation threshold. Theprogressive increase in crystallization rate in going from neat pellet PCLto extruded pellet PCL and with the further increase in the 1% and then5% CNC samples is clearly observed. (The 5% and 10% are similar, asdiscussed previously). The second trend is the shift in the percolationthreshold ξc that is exhibited in the extruded samples, either with orwithout added CNCs. In considering the origin of these two effects, wefirst consider the role of the CNCs (or impurities in the case of extrudedpellet PCL). As discussed earlier, it is known that CNCs and nano-particles in general can increase density of nucleation sites and thusincrease the crystallization rate. This effect is certainly responsible forthe increased crystallization rate that we observe here. However, it doesnot explain the decrease in the percolation threshold. Another way toincrease the nucleation density would be to decrease the crystallizationtemperature in an isothermal crystallization experiment. This was donein the previous work that modeled PCL crystallization via GEM, and itwas found that this threshold ξc is independent of temperature [50].Since ξc is reflective of the underlying geometry of the crystallizingentities, it is also intuitive that it should not be a function of nucleationdensity in the case where the nucleation density simply changes.

There is a limited temperature range over which these experimentscan be conducted. At lower values of Tc, the crystallization commencesbefore temperature equilibrium is achieved. At higher temperatures,the experiments become quite slow and difficult to reproduce. Thecrystallization kinetics of polycaprolactone (without filler) was mea-sured within the temperatures range of 40–44 °C by Kotula et al. [50].Though it is a limited temperature range, there were no qualitativechanges in the kinetics, just quantitative changes in rates. In the currentexperiments, we have found qualitatively similar behavior at 40 °C.

An explanation for the change in percolation threshold comes fromconsidering changes in the symmetry of crystallization. Nominally, we

expect that spherulites will grow in an isotropic fashion. However, if thecrystallization proceeds in an anisotropic fashion, then GEM predictsthat the percolation threshold will drop. As discussed in theIntroduction, it is known that nano-fibers can template polymer crys-tallization, so the resulting crystalline domains can be highly elongated.In the theoretical case of overlapping ellipsoids, the percolationthreshold is inversely proportional to the ellipse aspect ratio [48]. Wehypothesize that the crystals nucleated by the high aspect ratio CNCs,or even from small particles present from the extrusion process grow asanisotropic domains. If the impingement (percolation) occurs whilethese domains are still anisotropic, then a reduced percolationthreshold will be observed. The effect of the filler particles on thepercolation threshold is schematically illustrated in Fig. 9.

4. Conclusions

The modification of mechanical properties of bio-compatible semi-crystalline nanocomposites is complex because of the inter-related is-sues of dispersion, processing and crystallization phenomena.Controlling nanoparticle dispersion in polymer melts is a significantchallenge and measurement of its effect on crystallization is non-trivial.Here, we introduce a novel system for the study of a bio-derived na-noparticle in a biocompatible semi-crystalline polymer by modificationof a simple ion-exchange method. The simple ion exchange methodimproves the thermal stability while lowering the surface energies ofCNCs to allow for melt blending with hydrophobic polymers like PCL.We showed that the CNCs effectively act as nucleating agents enhancingthe overall crystallization rate and by analyzing the modulus-crystal-linity data using a generalized effective medium equation, we find acritical percolation threshold which decreases with the addition ofnanoparticles and extrusion. Overall, this work illustrates the com-plexity of crystallization phenomena in CNC nanocomposites and howvarious parameters including nanofiller surface area (greater surfacearea through modification of CNCs), percolation threshold and disper-sion quality in polymer matrix coordinate to govern the crystallizationkinetics. Making such connections between nanoparticle dispersion andorganization with macroscale properties is a crucial aspect for the op-timization of the processing conditions and the properties of the endproduct.

Fig. 9. Schematic showing crystallizationand percolation in a) pellet neat PCL b)pellet extruded PCL c) CNC filling samples.In neat PCL, the crystalline domains have tobe fairly large in order to overlap and per-colate. The crystalline fraction at which thishappens is higher. In the neat extruded PCL,small particles present due to extrusiongrow as anisotropic domains, as similar tothe CNC filled composite, which results insmaller percolation and the overall crystal-line fraction being lower for these samples.

D. Roy et al. Polymer 153 (2018) 70–77

76

Disclaimer

Certain commercial equipment, instruments, or materials are iden-tified in this paper in order to adequately specify experimental proce-dure. Such identification does not imply recommendation or endorse-ment by the National Institute of Standards and Technology, nor does itimply that the materials or equipment identified are necessarily the bestavailable for the purpose.

Acknowledgements

The authors wish to thank Wesley Farrell for performing GPCmeasurements.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.polymer.2018.08.007.

References

[1] H.-Y. Mi, X. Jing, J. Peng, M.R. Salick, X.-F. Peng, L.-S. Turng, Poly(ε-caprolactone)(PCL)/cellulose nano-crystal (CNC) nanocomposites and foams, Cellulose 21 (2014)2727–2741.

[2] C. Xiang, A.G. Taylor, J.P. Hinestroza, M.W. Frey, Controlled release of nonioniccompounds from poly(lactic acid)/cellulose nanocrystal nanocomposite fibers, J.Appl. Polym. Sci. 127 (2013) 79–86.

[3] J. Yang, J.-J. Zhao, C.-R. Han, J.-F. Duan, F. Xu, R.-C. Sun, Tough nanocompositehydrogels from cellulose nanocrystals/poly(acrylamide) clusters: influence of thecharge density, aspect ratio and surface coating with PEG, Cellulose 21 (2014)541–551.

[4] J. Juntaro, S. Ummartyotin, M. Sain, H. Manuspiya, Bacterial cellulose reinforcedpolyurethane-based resin nanocomposite: a study of how ethanol and processingpressure affect physical, mechanical and dielectric properties, Carbohydr. Polym.87 (2012) 2464–2469.

[5] A. Kumar, Y.S. Negi, N.K. Bhardwaj, V. Choudhary, Synthesis and characterizationof cellulose nanocrystals/PVA based bionanocomposite, Adv. Mater. Lett. 4 (2013)626–631.

[6] S. Sain, S. Sengupta, A. Kar, A. Mukhopadhyay, S. Sengupta, T. Kar, D. Ray, Effect ofmodified cellulose fibres on the biodegradation behaviour of in-situ formed PMMA/cellulose composites in soil environment: isolation and identification of the com-posite degrading fungus, Polym. Degrad. Stabil. 99 (2014) 156–165.

[7] V. Khoshkava, M.R. Kamal, Effect of Cellulose Nanocrystals (CNC) Particle mor-phology on dispersion and rheological and mechanical properties of polypropylene/CNC nanocomposites, ACS Appl. Mater. Interfaces 6 (2014) 8146–8157.

[8] E.H. Qua, P.R. Hornsby, Preparation and characterisation of nanocellulose re-inforced polyamide-6, Plast. Rubber Compos. 40 (2011) 300–306.

[9] A. Junior de Menezes, G. Siqueira, A.A.S. Curvelo, A. Dufresne, Extrusion andcharacterization of functionalized cellulose whiskers reinforced polyethylene na-nocomposites, Polymer 50 (2009) 4552–4563.

[10] C. Tian, S.Y. Fu, Q.J. Meng, L.A. Lucia, New insights into the material chemistry ofpolycaprolactone-grafted cellulose nanofibrils/polyurethane nanocomposites,Cellulose 23 (2016) 2457–2473.

[11] Y. Wang, C. Xu, D. Wu, W. Xie, K. Wang, Q. Xia, H. Yang, Rheology of the cellulosenanocrystals filled poly (є-caprolactone ) biocomposites, Polymer 140 (2018)167–168.

[12] T. Zimmermann, E. Pöhler, T. Geiger, Cellulose fibrils for polymer reinforcement,Adv. Eng. Mater. 6 (2004) 754–761.

[13] G. Siqueira, J. Bras, A. Dufresne, Cellulose bionanocomposites: a review of pre-paration, properties and applications, Polymers 2 (2010) 728–765.

[14] M.A.S. Azizi Samir, F. Alloin, A. Dufresne, Review of recent research into cellulosicwhiskers, their properties and their application in nanocomposite field,Biomacromolecules 6 (2) (2005) 612–626.

[15] P. Wambua, J. Ivens, I. Verpoest, Natural fibres: can they replace glass in fibrereinforced plastics, Compos. Sci. Technol. 63 (2003) 1259–1264.

[16] K. Oksman, A.P. Mathew, D. Bondeson, I. Kvien, Manufacturing process of cellulosewhiskers/polylactic acid nanocomposites, Compos. Sci. Technol. 66 (2006)2776–2784.

[17] M.A. Woodruff, D.W. Hutmacher, The return of a forgotten polymer-Polycaprolactone in the 21st century, Prog. Polym. Sci. 35 (2010) 1217–1256.

[18] V.R. Sinha, K. Bansal, R. Kaushik, R. Kumria, A. Trehan, Poly-є-Caprolactone mi-crospheres and nanospheres:an overview, Int. J. Pharm. 278 (2004) 1–23.

[19] F.P.W. Melchels, Marco A.N. Domingos, Travis J. Klein, J. Malda, P.J. Bartolo,D.W. Hutmacher, Additive manufacturing of tissues and organs 37 (2012)1079–1104.

[20] T. Zhou, J.-W. Zha, Y. Hou, D. Wang, J. Zhao, Z.-M. Dang, Surface-functionalizedMWNTs with emeraldine base: preparation and improving dielectric properties of

polymer nanocomposites, ACS Appl. Mater. Interfaces 3 (2011) 4557–4560.[21] D.M. Fox, R.S. Rodriguez, M.N. Devilbiss, J. Woodcock, C.S. Davis, R. Sinko,

S. Keten, J.W. Gilman, Simultaneously tailoring surface energies and thermal sta-bilities of cellulose nanocrystals using ion exchange: effects on polymer compositeproperties for transportation, infrastructure, and renewable energy applications,ACS Appl. Mater. Interfaces 8 (2016) 27270–27281.

[22] Y.-L. Loo, R.A. Register, A.J. Ryan, G.T. Dee, Polymer crystallization confined inone, two, or three dimensions, Macromolecules 34 (2001) 8968–8977.

[23] Y.-L. Loo, R.A. Register, A.J. Ryan, Modes of Crystallization in block copolymermicrodomains: breakout, templated, and confined, Macromolecules 35 (2002)2365–2374.

[24] C.M. Chan, J. Wu, J.X. Li, Y.K. Cheung, Polypropylene/calcium CarbonateNanocomposites, 43 (2002), pp. 2981–2992.

[25] J. Varga, I. Mudra, G.W. Ehrenstein, Highly Active Thermally Stable Β-nucleatingAgents for Isotactic Polypropylene, 74 (1999), pp. 2357–2368.

[26] T.A. Shephard, C.R. Delsorbo, R.M. Louth, J.L. Walborn, D.A. Norman, N.G. Harvey,R.J. Spontak, Self-Organization and Polyolefin nucleation efficiency of 1,3:2,4-Di-p-methylbenzylidene sorbitrol, J. Polym. Sci. B Polym. Phys. 16 (1997) 2617–2628.

[27] E. Di Maio, S. Iannace, L. Sorrentino, L. Nicolais, Isothermal crystallization in PCL/clay nanocomposites investigated with thermal and rheometric methods, Polymer45 (2004) 8893–8900.

[28] G. Lamberti, G.W.M. Peters, G. Titomanlio, Crystallinity and linear rheologicalproperties of polymers, Int. Polym. Process. 22 (2007) 303–310.

[29] J.-W. Housmans, R.J.A. Steenbakkers, P.C. Roozemond, G.W.M. Peters,H.E.H. Meijer, Saturation of pointlike nuclei and the transition to oriented struc-tures in flow-induced crystallization of isotactic polypropylene, Macromolecules 42(2009) 5728–5740.

[30] P.C. Roozemond, V. Janssens, P.V. Puyvelde, G.W.M. Peters, Suspension-likehardening behavior of HDPE and time-hardening superposition, Rheol. Acta 51(2012) 97–109.

[31] N.V. Pogodina, H.H. Winter, Polypropylene crystallization as a physical gelationprocess, Macromolecules 31 (1998) 8164–8172.

[32] K.B. Hafsia, M. Ponçot, D. Chapron, I. Royaud, A. Dahoun, P. Bourson, A novelapproach to study the isothermal and non-isothermal crystallization kinetics of Poly(Ethylene Terephthalate) by Raman spectroscopy, J. Polym. Res. 23 (2016) 1–14.

[33] Y. Hu, J.K. Liang, A.S. Myerson, L.S. Taylor, Crystallization monitoring by Ramanspectroscopy: simultaneous measurement of desupersaturation profile and poly-morphic form in flufenamic acid systems, Ind. Eng. Chem. Res. 44 (2005)1233–1240.

[34] M. Amer, Raman Spectroscopy for Soft Matter Applications, John Wiley & Sons,2009.

[35] A. Kelarakis, S.M. Mai, C. Booth, A.J. Ryan, Can rheometry measure crystallizationkinetics? A comparative study using block copolymers, Polymer 46 (2009)2739–2747.

[36] A.P. Kotula, M.W. Meyer, F. De Vito, J. Plog, A.R. Hight Walker, K.B. Migler, Therheo-Raman microscope: simultaneous chemical, conformational, mechanical, andmicrostructural measures of soft materials, Rev. Sci. Instrum. 87 (2016) 105105.

[37] S. Zhang, M.L. Minus, L. Zhu, C.-P. Wong, S. Kumar, Polymer transcrystallinityinduced by Carbon nanotubes, Polymer 49 (2008) 1356–1364.

[38] K. Lu, N. Grossiord, C. E. Koning, H. E.Miltner, B. V. Mele, J. Loos, Carbon nano-tube/isotactic Polypropylene composites prepared by latex technology: morphologyanalysis of CNT-induced nucleation, Macromolecules, 41 (2208) 8081-8085.

[39] L. Li, C.Y. Li, C. Ni, Polymer crystallization-driven, periodic patterning on Carbonnanotubes, J. Am. Chem. Soc. 128 (2006) 1692–1699.

[40] B. Natarajan, N. Lachman, T. Lam, D. Jacobs, C. Long, M. Zhao, B.A. Wardle,R. Sharma, J.A. Liddle, The evolution of carbon nanotube network sturcture inunidirectional nanocomposites resolved by quantitative electron tomography, ACSNano 9 (2015) 6050–6058.

[41] A.E. Lewandoska, S.J. Eichhorn, Quantification of the degree of mixing of cellulosenanocrystals in the thermoplastics using Raman spectroscopy, J. Raman Spectrosc.47 (2016) 1337–1342.

[42] A.P. Kotula, C.R. Snyder, K.B. Migler, Determining conformational order andcrystallinity in polycaprolactone via Raman spectroscopy, Polymer 117 (2017)1–10.

[43] Y. Wang, M. R-Perez, R.L. Reis, J.F. Mano, Thermal and thermomechanical beha-viour of polycaprolactone and starch/polycaprolactone blends for biomedical ap-plications macromol, Mater. Eng. 290 (2005) 792–801.

[44] K.B. Migler, A.P. Kotula, A.R. Hight Walker, Trans-Rich Structures in early stagecrystallization of polyethylene, Macromolecules 48 (2015) 4555–4561.

[45] J.S. Tiang, J.M. Dealy, Shear -induced crystallization of isotactic polypropylenestudied by simultaneous light intensity and rheological measurements, Polym. Eng.Sci. 52 (2012) 835–848.

[46] Z. Xiao, Y.A. Akpalu, New insights into the characteristics of early stage crystal-lization of a polyethylene, Polymer 48 (2007) 5388–5397.

[47] B. Heck, T. Kawai, G. Strobl, Time dependent light attenuation measurements usedin studies of the kinetics of polymer crystallization, Polymer 47 (2006) 5538–5543.

[48] S. Cotugno, E. Di Maio, C. Ciardiello, S. Iannace, G. Mensitieri, L. Nicolais, Sorptionthermodynamics and mutual diffusivity of carbon dioxide in molten poly-caprolactone, Ind. Eng. Chem. Res. 42 (2003) 4398–4405.

[49] E.J. Garboczi, K.A. Snyder, J.F. Douglas, M.F. Thorpe, Geometrical percolationthreshold of overlapping ellipsoids, Phys. Rev. 52 (1995) 819–828.

[50] A.P. Kotula, K.B. Migler, Evaluating models for polycaprolactone crystallization viasimulatneous rheology and Raman spectroscopy, J. Rheol. 62 (2018) 342–356.

D. Roy et al. Polymer 153 (2018) 70–77

77